The disclosed computer-implemented method may include identifying data format requirements for one or more machine-learning-based audio/video classifiers. The classifiers may be configured to detect classifiable features of decoded audio/video data. The method may also include decoding once, for the one or more classifiers, a video stream into audio/video data based on the identified data format requirements, and creating a new instance of each of the one or more classifiers. In addition, the method may include transforming the audio/video data for each instance based on the respective data format requirements and providing the respective transformed audio/video data to each instance. The method may also include performing, in parallel, classification of each transformed audio/video data by each respective instance and then terminating each instance of the one or more classifiers. Various other methods, systems, and computer-readable media are also disclosed.
Legal claims defining the scope of protection, as filed with the USPTO.
1. A computer-implemented method comprising: identifying data format requirements for one or more machine-learning-based audio/video classifiers, the classifiers being configured to detect classifiable features of decoded audio/video data; decoding once, for the one or more classifiers, a video stream into audio/video data based on the identified data format requirements; creating a new instance of each of the one or more classifiers; transforming the audio/video data for each instance based on its respective data format requirements; providing the respective transformed audio/video data to each instance; performing, in parallel, classification of each transformed audio/video data by each respective instance; and terminating each instance of the one or more classifiers.
2. The method of claim 1 , further comprising updating at least one of the one or more classifiers independently from instances of the one or more classifiers.
3. The method of claim 2 , wherein the new instances of each of the one or more classifiers are created from the updated versions of the one or more classifiers.
4. The method of claim 2 , wherein the data format requirements are identified from the updated versions of the one or more classifiers.
5. The method of claim 1 , wherein transforming the audio/video data further comprises downsampling the audio/video data for at least one of the instances of the one or more classifiers when the audio/video data exceeds the data format requirements for the at least one instance.
6. The method of claim 1 , wherein transforming the audio/video data further comprises cropping the audio/video data for at least one of the instances of the one or more classifiers to satisfy the data format requirements for the at least one instance.
7. The method of claim 1 , wherein decoding the video stream further comprises decoding the video data into a lowest quality video data satisfying the data format requirements.
8. A system comprising: one or more processors; a memory device comprising instructions that, when executed by the one or more processors, cause the system to: identify data format requirements for one or more machine-learning-based audio/video classifiers, the classifiers being configured to detect classifiable features of decoded audio/video data; decode once, for the one or more classifiers, a video stream into audio/video data based on the identified data format requirements; create a new instance of each of the one or more classifiers; transform the audio/video data for each instance based on its respective data format requirements; provide the respective transformed audio/video data to each instance; perform, in parallel, classification of each transformed audio/video data by each respective instance; and terminate each instance of the one or more classifiers.
9. The system of claim 8 , wherein the instructions further comprise instructions for updating at least one of the one or more classifiers independently from instances of the one or more classifiers, wherein the new instances of each of the one or more classifiers are created from updated versions of the one or more classifiers.
10. The system of claim 9 , wherein the data format requirements are identified from updated versions of the one or more classifiers.
11. The system of claim 8 , wherein transforming the audio/video data further comprises downsampling the audio/video data for at least one of the instances of the one or more classifiers when the audio/video data exceeds the data format requirements for the at least one instance.
12. The system of claim 8 , wherein transforming the audio/video data further comprises cropping the audio/video data for at least one of the instances of the one or more classifiers to satisfy the data format requirements for the at least one instance.
13. The system of claim 8 , wherein decoding the video stream further comprises decoding the video data into a lowest quality video data satisfying the data format requirements.
14. A non-transitory computer-readable medium comprising computer-readable instructions that, when executed by at least one processor of a computing device, cause the computing device to: identify data format requirements for one or more machine-learning-based audio/video classifiers, the classifiers being configured to detect classifiable features of decoded audio/video data; decode once, for the one or more classifiers, a video stream into audio/video data based on the identified data format requirements; create a new instance of each of the one or more classifiers; transform the audio/video data for each instance based on its respective data format requirements; provide the respective transformed audio/video data to each instance; perform, in parallel, classification of each transformed audio/video data by each respective instance; and terminate each instance of the one or more classifiers.
15. The non-transitory computer-readable medium of claim 14 , further comprising instructions for updating at least one of the one or more classifiers independently from instances of the one or more classifiers.
16. The non-transitory computer-readable medium of claim 15 , wherein the new instances of each of the one or more classifiers are created from updated versions of the one or more classifiers.
17. The non-transitory computer-readable medium of claim 15 , wherein the data format requirements are identified from updated versions of the one or more classifiers.
18. The non-transitory computer-readable medium of claim 14 , wherein transforming the audio/video data further comprises downsampling the audio/video data for at least one of the instances of the one or more classifiers when the audio/video data exceeds the data format requirements for the at least one instance.
19. The non-transitory computer-readable medium of claim 14 , wherein transforming the audio/video data further comprises cropping the audio/video data for at least one of the instances of the one or more classifiers to satisfy the data format requirements for the at least one instance.
20. The non-transitory computer-readable medium of claim 14 , wherein decoding the video stream further comprises decoding the video data into a lowest quality video data satisfying the data format requirements.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
February 23, 2018
June 9, 2020
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.